Keywords

In the previous chapter, we laid down the conceptual foundations of our approach towards the relationship between media and illiberalism. We outlined how media play a key role in the rise of the illiberal public sphere, facilitating its progression from the incipient to the ascending stage and potentially growing further into the hegemonic stage, as illustrated by the examples of Hungary and Serbia. So far, we have primarily discussed this phenomenon in relation to institutions, observing how news media are captured by illiberal actors and used as channels for the dissemination of an illiberal rhetoric and programmes. In this chapter, we gradually shift our focus from institutional settings and actors to media audiences. We examine audiences’ news consumption patterns, as well as the extent to which these patterns are mirrored by people’s attitudes towards selected polarizing issues that are frequently exploited by illiberal actors across the four countries we examine.

Drawing on our expert survey, we begin our analysis by mapping out the editorial bias of a sample of major news media in each of the four countries. We do so to ascertain the degree of polarization that characterizes these countries’ media systems and its relation to ideological and political axes. Utilizing data from the population survey, we then explore and compare audiences’ news media diets—or repertoires—which we assess in relation to their political and ideological orientation and diversity. The central part of the chapter is devoted to the investigation of the relationship between these patterns of news consumption and people’s voting preferences, as well as (il)liberal attitudes. Our analysis foregrounds two dimensions of illiberalism: cultural (represented particularly by attitudes towards minorities and their rights) and constitutional (i.e. the importance attributed to selected democratic institutions and principles). Beyond news repertoires, we also specifically consider the link between (il)liberal attitudes and exposure to particular types of media, namely public service broadcasters, commercial television, and radio. We conclude by interpreting our findings within the broader context of the evolution of the illiberal public sphere in the four examined countries.

3.1 Media Polarization: Whither the Centre?

The rise of illiberalism in democratic countries is usually accompanied by a deepening polarization of both political and media landscapes alongside the liberal/illiberal axis. Defined by McCoy et al. (2018) as ‘a process whereby the normal multiplicity of differences in a society increasingly aligns along a single dimension, cross-cutting differences become reinforcing, and people increasingly perceive and describe politics and society in terms of “us” versus “them”’ (p. 18), polarization can serve as a powerful political strategy to mobilize the electorate by exploiting existing societal cleavages (McCoy & Somer, 2021). While the use of such a strategy is not necessarily limited to illiberal parties and politicians, they nonetheless are the ones who tend to adopt it most frequently and benefit from it. According to Somer and McCoy (2018), this is because polarization ‘advantages actors willing and able to employ unyielding, exclusionary, and demagogic politics and rhetoric’, while facilitating ‘the development of rigid and antagonistic political identities’ (p. 5). Moreover, polarization also works in favour of illiberalism by severely restricting the common ground where a compromise might be reached, weakening democratic norms and eroding trust in institutions that are supposed to serve the entire society (McCoy & Somer, 2021; Svolik, 2019; Carothers & O’Donohue, 2019).

While both elite polarization (i.e. the divergences between political parties) and mass polarization (i.e. the cleavages in public attitudes on political issues) have been subject to an increasing amount of research, especially in recent years (Moral & Best, 2023; Zingher, 2022; Enders, 2021; Diermeier & Li, 2019), empirical explorations of media polarization are still relatively scarce, particularly comparative ones. The question of how far apart news media are when it comes to their ideological leaning or political orientation has been traditionally examined within the conceptual framework of media partisanship or bias, and empirically investigated using different methods, most commonly content analysis (Shultziner & Stukalin, 2021; Hameleers, 2019; Budak et al., 2016), expert surveys (Popescu et al., 2011; Castro-Herrero et al., 2016), or a combination of the two (Castro, 2021). However, available studies on media polarization in Eastern Europe are either substantially outdated (Castro-Herrero et al., 2016) or limited with regard to their geographical reach or variety of news outlets (Olechowska, 2022).

Filling this gap has therefore been the necessary first step to lay the foundations for our analysis of the relationship between audience news choices and their adherence to illiberal values. Focusing on the ideological orientation (i.e. conservative vs. liberal bias) and general political leaning (measured as pro- vs. anti-government editorial stance) of selected news brands, our mapping was carried out by means of an expert survey that assessed a sample of the most politically relevant news outlets in each country (for more details on the methodology of the expert survey, as well as on the selection individual news brands, see the Methodological Appendix at the end of this book). Given that the data were collected in spring 2020, the assessment of both the political and ideological biases of the selected outlets, and, consequently, of the polarization of the news media system as a whole, reflects the situation at that specific time point. However, because we are primarily interested in determining the depth of political and ideological divides across news media systems and in exploring the relationship between media and audience polarization, the mapping is relevant regardless of any changes that might have happened since the collection of the data.

The results (Fig. 3.1) paint a picture of strongly polarized media landscapes in Hungary and Serbia where there is a clear divide between news media brands that display a pro-government bias and those whose editorial line is critical towards the government, with none positioned around the centre of this scale. In Serbia, most of the news outlets represent the former camp, including the brands with the biggest audience reach, such as the tabloid Blic, or the commercial TV stations ALO or Pink. The anti-government media bloc is composed of a handful of low-circulation papers—the daily Danas, the weekly Vreme, the 24/7 cable news channel N1 (an affiliate of CNN, operating from Slovenia), and the Balkan branch of the US-funded Radio Free Europe (Radio Slobodna Evropa, RFE), broadcasting for the former Yugoslav countries. The comparison between these two camps makes it clear which one significantly prevails. Therefore, while we can still describe the Serbian media landscape as being polarized, we must also recognize that this polarization is heavily asymmetrical, leaning in favour of the government.

Fig. 3.1
A dot plot of anti or pro government stance has data for C Z E, H U N, P O L, and S R B. The graph is divided into 3 sections, anti government, neutral, and pro government. P O L and H U N, most of the dots fall in anti government. S R B and C Z E, most dots are in pro government.

Political bias of selected news outlets

Full size image

The extent of polarization in Hungary is similar. However, in contrast to Serbia, some of the anti-government brands have a large following, especially the commercial television channel RTL Klub, the news websites 444.hu and Index.hu (since the time of our survey, the latter has changed ownership and editorial line, becoming a Fidesz supporter and thereby adopting a pro-government position).Footnote 1 In addition, compared with Serbia, the pro-government media bloc is much more closely aligned in terms of its level of support for the government, indicating closer political control over these outlets.

The other two countries—the Czech Republic and Poland—display a greater variety of political bias across selected outlets, implying a relatively lower level of media polarization. The key outlets in the pro-government media bloc are the state-controlled public service media (TVP, Polish Radio), the online outlet wPolityce.pl—whose ownership is crucially linked to the Law and Justice party—and the catholic Radio Maryja, infamous for its ultra-conservative agenda (Krzemiński, 2016). On the other side of the spectrum, there are outlets belonging to the main oppositional media house Agora—the biggest non-tabloid daily Gazeta Wyborcza and the radio station TOK FM—as well as US-owned broadcaster TVN, a long-term target of government attacks and pressure. However, there are also several news brands which that can be classified as belonging to a ‘neutral’ zone or close to it, including the biggest tabloid Fakt and the largest commercial television network Polsat.

The Czech media landscape appears to be the least polarized of all four countries, with individual brands spread across the scale rather evenly. Even though some of them might have swapped their anti- and pro-government orientation since data collection—following the 2021 parliamentary elections which installed a new, non-populist government—none of them is positioned as far away from the centre as some of the brands in Poland, Hungary, and Serbia. Finally, there are multiple brands that occupy the ‘neutral’ zone, including public service media (Czech Television and Czech Radio), as well as the biggest tabloid Blesk.

The expert ranking of news media brands by their orientation on the liberal-conservative scale (see Fig. 3.2) complements the previous map of the brands’ political bias. Again, Hungary and Serbia are the more polarized ones, with a significant gap between the liberal and conservative camps and no outlet in the ‘neutral’ area. The Polish and Czech media landscapes are characterized by relatively greater ideological diversity, with the Czech Republic lacking in outlets that would be positioned on the far ends of the spectrum.

Fig. 3.2
A dot plot of ideology has data for C Z E, H U N, P O L, and S R B. The graph is divided into 3 sections, liberal, neutral, and conservative. P O L and H U N, most of the dots fall in liberal. S R B and C Z E, most dots are in conservative.

Ideological bias of selected news outlets

Full size image

Overall, the combination of these two maps, that—as formerly clarified—reflect the situation in the first half of 2020, suggests that media polarization is more deeply engrained in the countries with the most advanced illiberal public sphere, Hungary and Serbia. Whether assessed by virtue of their support for the government or by their ideological inclination (two elements that, for most brands, strongly correlate), the media in these two countries are sharply divided, with the pro-government/conservative camp visibly dominating in Serbia in terms of its market strength. The Czech media system appears to be the only one withstanding the centrifugal tendencies observed in the other countries, an aspect that might be attributed to the strong position that public service media occupy in the country’s media landscape. The fact that both public service radio and television are currently leaders in their respective market segments and simultaneously enjoy a high level of trust stemming from the perceived quality of their output, as well as editorial independence (Newman et al., 2021), means that they are in an influential position to set the standards for the rest of news media organizations, preventing the country from spiralling down the process of partisanship and polarization observed in the other countries.

Fig. 3.3
A scatterplot of negative and positive values for anti or pro government stance versus ideology liberal or conservative has dots for Czech Republic, Hungary, Poland, and Serbia. The dots fall in a diagonally increasing pattern from negative to positive values.

Map of political—ideological bias of selected news outlets

Full size image

To further demonstrate the extent of polarization across the region, the two previously presented graphs are plotted onto a 2-D map, showing the position of each of the news brands simultaneously on both axes (Fig. 3.3). As it is apparent, only two of the four quadrants are populated: the bottom-left one (comprising brands that are both anti-government and liberal-oriented), and the top-right one (occupied by brands that display both pro-government and conservative bias). In other words, at the time of mapping, there were virtually no (significant) news media outlets in either of the four countries that would be classified as both pro-government and liberal, or anti-government and conservative. Even though this is arguably no longer the case in the Czech Republic, where some of the conservative media adopted a government-critical editorial stance since the October 2021 elections (most notably the dailies controlled by the ex-PM Andrej Babiš), the map represents a plastic portrayal of the overall diminishing plurality within the media system in Eastern Europe as a consequence of growing political polarization and media partisanship.

3.2 Mapping Liberal Attitudes: Cultural and Constitutional Liberalism

Having outlined the extent of polarization across the media landscapes of the four examined countries, based on the expert assessment of the political and ideological bias of selected news brands, we now turn to audiences to examine the extent to which their news consumption habits reflect this polarization. Our assessment is aimed at establishing whether audiences’ exposure to particular news brands aligns with their political attitudes, especially on issues that are known to be polarizing in those countries and are often utilized by illiberal actors to mobilize their electorate.

Drawing on the population survey data collected between December 2019 and January 2020, we start by outlining the prevalent attitudes we observed among audiences. To empirically gauge people’s attitudinal susceptibility towards illiberalism, we have constructed two composite measures derived from a set of survey questions. The first one, labelled cultural liberalism index, is based on the responses to eleven individual questions, inquiring on the respondents’ opinions on immigration (i.e. whether immigration poses a threat to local culture, causes the rise of criminality or abuse of the welfare system), same-sex marriage and same-sex adoption, and assessing, more broadly, their tolerance towards specific minorities, particularly immigrants, same-sex couples, Muslims, Roma, Jewish and black people (see the Electronic Supplementary Material for the exact wording of the questions). The second measure is what we call constitutional liberalism index, which quantifies the importance that respondents assign to some of the key institutions underlying modern-day liberal democracy. These are: (a) free and fair elections, (b) law and order, (c) freedom of speech, (d) peace and stability, (e) independent media, (f) strong political opposition, (g) a courts system that treats all citizens equally, and (h) equal rights for women (Table 3.1).Footnote 2

Table 3.1 Cultural and constitutional liberalism index
Full size table

The results show far greater variance on the cultural liberalism index than on the constitutional liberalism index. The cultural liberalism index largely follows a normal distribution (see the Electronic Supplementary Material), with Poland displaying the highest average score of the four countries (4.06 on a 1–7 scale), followed by the Czech Republic (3.96), Hungary (3.89) and Serbia (3.73). The fact that the Polish respondents appear to be more ‘liberal’ than Czech respondents—contrary to what could be expected based on the comparative state of the illiberal public sphere in the two countries, which is more advanced in Poland—is explained by differences between these countries on the specific dimensions of the index. While the Czech respondents show relatively high support of same-sex marriage (60% in favour, compared with Poland at 38.5%) and same-sex adoption (55% in favour, compared with 22% in Poland), they are significantly more hostile towards immigration (64% agreeing that ‘immigrants are a threat to our culture’ and 66% that ‘immigrants cause the rise of criminality’, compared with 43% and 46.5% in Poland).

In contrast to the distribution of scores on the cultural liberalism index, the index of constitutional liberalism is heavily skewed towards the top end, with the average scores (on a 1–7 scale) showing very small gaps between the four countries (Serbia 6.43, Poland 6.41, Hungary 6.40, the Czech Republic 6.22). This reveals that respondents agree on the chief importance of institutions of liberal democracy, even at a time when they are being undermined by illiberal actors across the region. Nevertheless, it is still possible to test whether attitudes towards democratic institutions are impacted by their news media exposure. This analysis will be the focus of the following section.

3.3 From Individual News Brands to Media Repertoires: Selective Exposure as a Cross-Media Phenomenon

Before getting into the data on the relationship between news consumption and political attitudes, it is first necessary to explain the approach we have used for measuring what kind of news media people are exposed to. While our research draws on selective exposure theory (Lazarsfeld et al., 1948; Klapper, 1960; Bennett & Iyengar, 2008), which argues that people prefer to be exposed to messages that are congruent with their existing opinions and beliefs rather than those that challenge them, we do not follow the traditional approach towards testing it. Instead of observing and comparing people’s exposure to individual news brands in isolation from each other (e.g. Iyengar & Hahn, 2009; Gil de Zúñiga et al., 2012; Arendt et al., 2019), we have based our analysis on the innovative concept of media repertoires (Hasebrink & Domeyer, 2012; Hasebrink & Hepp, 2017). We have chosen this methodology because we believe it better reflects the nature of today’s high-choice media environment where people’s news media diets are likely to be composed of a variety of sources comprising different media types, rather than a singular medium type (i.e. television, newspapers, online platforms) or even a single channel (see Tóth et al., 2023, for a more detailed elaboration of this approach).

In recent years, this method has been increasingly employed to measure news consumption and analyse behavioural and attitudinal variables such as political participation (Strömbäck et al., 2018), political interest and knowledge (Kim, 2016), and partisan preferences (Edgerly, 2015), or to quantify support for political actors, such as Donald Trump (Mourão et al., 2018). However, our approach deviates from much of these previous studies, as we focus on the underlying political and ideological position of the individual media brands consumed by our respondents, instead of simply grouping together media brands by type or genre. This means that the repertoires are indicative of the relative political/ideological homogeneity or heterogeneity of people’s news media diets. In other words, they show how resistant or open people are to information and opinions from across the political and ideological spectrum.

To ascertain the number and character of news media repertoires across our survey sample, we have first identified respondents’ ‘regular’ media consumption (i.e. at least once a week) of the aforementioned pre-selected news media brands, spanning a variety of media types (Fig. 3.1 and Fig. 3.2; also see Methodological Appendix for their full list). Utilizing the scores for ideological and political bias of such news brands as evaluated by our expert survey, we then used latent profile analysis (LPA) to calculate the probability with which the respondents’ news consumption patterns fall into specific, ideologically and politically coherent, media repertoires.Footnote 3

The analysis revealed five different types of media repertoires, distinguished by their level of homogeneity or heterogeneity (in other words, ‘closedness’ or ‘openness’) with regard to the political-ideological position of the news brands consumed by the respondents. The repertoires are broadly positioned alongside the diagonal axis cutting across the four quadrants in Fig. 3.3, dividing the media map between anti-government/pro-liberal brands on the one hand side, and pro-government/conservative brands on the other. Their classification as ‘open’ or ‘closed’ is determined by the extent to which respondents—while showing preference for one or the other side—also allow into their news media diets some of the brands from the opposite camp. Hence, what we define as the ‘closed liberal anti-government repertoire’ and the ‘closed conservative pro-government repertoire’ are both characterized by politically and ideologically very coherent media diets, with only a minimal presence of the sources representing a notably divergent editorial line. Conversely, people who fall within the ‘open liberal anti-government repertoire’ and the ‘open conservative pro-government repertoire’ are, as the names suggests, relatively less orthodox when it comes to the editorial leaning of their chosen news brands, even though, overall, one side still clearly prevails over the other. Finally, the ‘balanced media repertoire’ comprises people whose media consumption is most diverse and non-discriminant, showing a slight preference for politically and ideologically neutral media, but still exposed to plenty of liberal and conservative media and, to a lesser extent, pro-government brands.

The proportions of each of these five repertoires in the four countries are displayed in Fig. 3.4. As we can see, with the exception of Serbia, ‘open’ repertoires are more common than ‘closed’ ones, suggesting that extreme forms of selective exposure might be relatively less prominent than commonly assumed. In the Czech Republic, the amount of people whose news diets can be said to form a kind of ideological/political ‘bubble’ (on both sides of the spectrum) is less than 17% and, in Poland, it is around 20%. In Hungary, however, it is more than a third of the sample (34%) and, in Serbia, the same figure rises to nearly 50%. These two countries are also characterized by a complete absence of the fifth—balanced—repertoire, in line with the previously observed gap in the political and ideological ‘centre’ on their respective media maps. The sizeable proportion of the balanced repertoire in the Czech Republic (31%) is significantly aided by the strong market position of public service television and radio, both demonstrably neutral in their political leanings.

Fig. 3.4
A horizontal graph of 4 countries versus percent of respondents has bars indicating the proportions of 5 types of governments, closed liberal anti government, open conservative pro government, open liberal anti government, closed conservative pro government, and balanced. S R B has a maximum of closed conservative pro government.

The five media repertoires

Full size image

3.4 Media Repertoires and Liberal Attitudes

To finally address the question of whether and how people’s news consumption is reflected in their political preferences and attitudes, this chapter will now turn to the examination of the relationship between the key variables considered in our survey, namely people’s news repertoires and the indexes of cultural and constitutional liberalism.

Figure 3.5 shows the breakdown of the average scores on the cultural liberalism index (1–7 scale) across the five media repertoires in each country. In three of the four countries, people who consume news almost exclusively from liberal anti-government media (i.e. have a closed liberal anti-government news repertoire) are culturally the most liberal. The only exception is the Czech Republic, where people with an open liberal anti-government repertoire are slightly more culturally liberal than those with a closed one. However, the difference between the two means is not significant because of the low N in both groups, as apparent from the Fig. 3.5 (in which the sizes of the individual ‘bubbles’ represent the proportion of people characterized by the specific media repertoire in each country). Conversely, people whose news diets are restricted merely to conservative pro-government sources exhibit the least liberal attitudes. This is the case in all four countries, with the lowest scores on the index being displayed by conservative pro-government news consumers in Hungary (3.08), followed by Poland (3.23), the Czech Republic (3.48) and Serbia (3.49). Overall, the results confirm that, in line with selective exposure theory, a link exists between the kind of media people are exposed to and their attitudes towards selected cultural issues that have a potential to polarize the public alongside the liberal/illiberal axis in these countries. Across our sample, people whose news diets are characterized by consuming primarily news brands with a liberal editorial orientation (which, at the time of data collection, also overlapped with these brands’ anti-government profile), tend to display relatively more liberal attitudes than those who predominantly consume conservative, pro-government sources.

Fig. 3.5
A dot plot of percentage of closed or open liberal anti government, open or closed conservative pro government, and balanced indicated by dots of various sizes, indicating percentage from 10% to greater than 60%. Poland has the maximum of greater than 60% of open liberal anti government.

Media repertoires and cultural liberalism

Full size image

As for the attitudes towards some of the key institutions of liberal democracy captured by the constitutional liberalism index, the differences in average scores among the individual media repertoire groups are less prominent (see the Electronic Supplementary Material for full data), reflecting the relatively low variance of this index, as discussed earlier in this chapter. However, they point in a similar direction, thereby corroborating the trend highlighted by the cultural liberalism index. Across all four countries, the people who attribute the highest importance to democratic institutions are typically those whose news media diets consist solely of liberal anti-government sources, while those who consume predominantly conservative pro-government sources are comparatively less persuaded of the significance of these institutions, though still being largely supportive of them.

Apart from demonstrating general associations between people’s selective news exposure and their attitudes, the data captured in Fig. 3.5 offer an additional opportunity to explore how close or far apart the members of individual repertoire groups are with regard to their views on selected cultural issues (as represented by the cultural liberalism index), thereby indicating the extent of polarization in each country. Comparing the average scores of respondents with closed liberal anti-government news diets and those with closed conservative pro-government ones, we can see that the gap between those two opposite groups is the widest in Hungary (1.89), and the narrowest in the Czech Republic (1.36), with Poland (1.57) and Serbia (1.45) in the middle.

Furthermore, comparing the levels of cultural liberalism across all four repertoire groups that display partisan bias (that is, leaving out the balanced repertoire, which is effectively present only in the Czech Republic), we can observe interesting patterns in the way these scores are distributed on the scale. In Poland, the difference between the average scores for the members of the first and second repertoire (closed- and open liberal anti-government) is virtually the same as the gap between the members of the second and fourth repertoire. These data indicate that people who have predominantly liberal and anti-government news diets but occasionally mix them with conservative and pro-government sources are, after all, not too significantly distant in their cultural attitudes from their counterparts on the right (those consuming predominantly conservative pro-government media, but sometimes being exposed to liberal content, too). In other words, the division between these two open repertoires in Poland is less sharp than in the Czech Republic, where the consumers of liberal anti-government media and the consumers of conservative pro-government outlets—regardless of whether they fall into an open or closed variant of the respective repertoires—are distinctly far apart from each other when it comes to their attitudes to selected cultural topics.

The same pattern is even more prominent in Hungary, where the average score on the cultural liberalism index for news consumers who fall within the open liberal repertoire is actually much nearer to the closed conservative repertoire than to the closed liberal one. This shows that the views of people in the open liberal repertoire on selected cultural issues are more likely to overlap with those displayed by the audiences of conservative outlets, rather than those of consumers who only stick to liberal anti-government sources. To put it more bluntly, in Hungary it is only the members of the first, closed liberal anti-government repertoire who stand out as culturally liberal, while the remaining three form a distinct cluster that, overall, shows adherence to more illiberal than liberal values. This puts the previously outlined shares of individual repertoires (Fig. 3.4) in a different perspective: even though the largest proportion of audiences in Hungary are characterized by open liberal anti-government news diets (51.3%), they do not appear to be particularly liberal-oriented, at least when it comes to their views on selected cultural issues.

The findings regarding the link between selective partisan exposure and attitudes on polarizing issues are further supported by the electoral choices of the members of each of the media repertoires across the four countries (see Table 3.2). Overall, the pattern observed for news consumers in both of the closed repertoires is similar to the one detected in relation to the cultural liberalism index. In all four countries, people predominantly getting their news from liberal anti-government sources had a statistically higher probability to have voted against the governing party in the last national elections (2016 in Serbia, 2017 in the Czech Republic, 2018 in Hungary, and 2019 in Poland), while citizens preferring conservative, pro-government news brands were more likely to have voted in support of the incumbent government. The findings are a bit less clear for those whose news diets have been more open. The open liberal anti-government repertoire only shows statistically significant correlations with voting against Aleksandar Vučić’s government in Serbia, while in the other countries, the results are statistically inconclusive. Similarly, the open conservative pro-government repertoire shows significant correlations with pro-government vote in Hungary (Fidesz) and Poland (Law and Justice), but not in the Czech Republic (ANO 2011) and Serbia (Serbian Progressive Party), though in both these countries, the relationships between these variables point in the expected direction. However, regardless of statistical significance, it is worth pointing out that the patterns found in Hungary virtually mirror those outlined in the previous paragraph: the only repertoire whose members were likely to vote against Viktor Orbán’s government in 2018 was the first (closed liberal anti-government) one, while the members of the second repertoire (open liberal anti-government) were clearly divided between pro- and anti-government parties, and overall closer to the voters whose news diets preferred conservative pro-government sources.

Table 3.2 Voting for/against the government party in the last elections by news repertoires
Full size table

3.5 Public Service Media: Channels of Illiberalism, or Harbours for Liberal Audiences?

So far, the relationship between media exposure and (il)liberal attitudes among Eastern European publics has been analysed with a focus on media repertoires, emphasizing a more complex perspective on news consumption patterns and on their role in shaping attitudes in the contemporary high-choice media environment. Aside from this theoretical reason, we also avoided exploring such links in the context of individual news brands for practical and statistical reasons, given the relatively low number of users of most news brands in our survey. Nevertheless, we shall now zoom in on the meso level of selected media types and look specifically at public service media (PSM)—a notoriously contested institution across the region, as described in more detail in Chap. 2.

The results of our analysis, captured by Table 3.3, reveal clear differences among the four examined countries when it comes to the cultural attitudes of PSM audiences. In the Czech Republic, watching Czech Television or listening to Czech Radio significantly increases the probability of leaning towards liberal values on the cultural liberalism index. Likewise, attitudes to liberal democratic institutions also strongly correlate with exposure to public service media broadcasters. This represents a stark contrast to Hungary, where audiences of PSM are presenting the opposite pattern: the more frequent exposure to MTV, TV2, or Kossuth Radio, the lower people score on both the cultural and constitutional liberalism indices. Similarly, in Poland, frequently watching TVP or listening to Polish Radio makes it more likely for people to display illiberal attitudes with regard to cultural issues. However, this is not the case for constitutional liberalism, where the effect of PSM exposure is not statistically significant. The opposite is true for frequent audiences of news programmes broadcast by Radio Television of Serbia (Radio-televizija Srbija, RTS) who do not show particular preferences on cultural issues (on the cultural liberalism index), but attribute significantly lower importance to democratic institutions (as measured by the constitutional liberalism index) than those who consume news from this broadcaster less frequently. Furthermore, in contrast to Poland and Hungary where heavier consumption of commercial television and radio does not seem to correlate with any preference for liberal or illiberal attitudes, Serbian consumers of commercial channels are more likely to hold illiberal attitudes on cultural issues than those who primarily consume public service channels. This suggests that, in Serbia, where the illiberal public sphere is most entrenched, both commercial and public service media contribute to the reproduction and spreading of illiberalism, albeit in different forms.

Table 3.3 Impact of consuming public service vs. commercial media news on liberal attitudes
Full size table

These more granular data largely confirm the general patterns outlined above, namely the congruence between exposure to media that display a particular political-ideological orientation and the corresponding views of their audiences. With regard to public service broadcasters, this overlap is clearest in Hungary, where the institution of PSM has been effectively turned into an instrument of government propaganda and is known to be promoting illiberal values (Tamássy, 2019; Kondor et al., 2022). A similar situation is observed in Poland and Serbia although, in both cases, the match between the exposure to PSM and people’s attitudes only relates to one of our two measures of attitudinal illiberalism. In Poland, it concerns attitudes towards culturally polarizing issues, actively exploited by the government-controlled Polish national broadcasters TVP and Polish Radio. In Serbia, it relates to the perceived significance of democratic institutions, as captured by the constitutional liberalism index. The liberal orientation of Czech Television and Czech Radio, which both enjoy a fair amount of political independence, is comparatively much less prominent than the conservative bias of their counterparts in the other three countries (see Fig. 3.2); however, their audiences still exhibit relatively strong liberal attitudes. Interestingly, the negative effect of consuming news from Czech commercial broadcasters on cultural liberalism is just as strong as the effect of consuming news from the Hungarian state-controlled PSM. This further supports the argument made in Chap. 2 that the channels of the illiberal public sphere are potentially not just limited to media captured by illiberal governments or political actors, but can include commercial media too, often simply because exploiting culturally polarizing issues reveals to be a profitable business strategy, as is well known from the United States and other Western countries (Klein, 2020).

3.6 Conclusions

This chapter has presented several key findings from the analysis of both the expert and population surveys which shed more light on the contemporary processes of media and political polarization across Eastern Europe, improving our understanding of how news consumption might be related to these trends. First, we have demonstrated that the news media landscapes in the four countries under our investigation have a tendency towards polarization alongside the political and ideological axis. Nevertheless, the actual level of polarization—the extent of the divide between the two opposite poles of the political spectrum—exhibits some variations across the sample. These variations broadly correspond to the different stages of the illiberal public sphere observed in the countries and outlined in Chap. 2. Hungary and Serbia, the two countries at the most advanced (hegemonic) stage, clearly display the highest degree of polarization in terms of the positioning of individual news brands across both the political and ideological spectrum. However, the data from our sample reveal that polarization in Serbia is also very asymmetrical, heavily tilted towards the pro-government and conservative side, which is represented by substantially more news brands and with notably bigger audience reach than those labelled as anti-government and pro-liberal. In Poland, where the illiberal public sphere has reached the ascendant but not yet hegemonic stage, the news landscape is comparatively more varied. However, some of the key representatives of the opposite camps—such as the government-controlled TVP and the leading quality newspaper Gazeta Wyborcza—are still positioned at the far sides of both the political and ideological spectrum. Finally, in the Czech Republic where the illiberal public appears comparatively the least developed, situated between the incipient and ascendant stages, the news system is the least polarized of all four countries, although still showing some clear divides, particularly along the ideological axis. Nevertheless, the Czech Republic also has a relatively sizeable ‘centre’, represented by the presence of several brands that are classified as relatively neutral and impartial (especially the public service media), which we believe acts as a countervailing force against the centrifugal pressures of polarization.

We have subsequently explored audiences’ news media diets and classified them into five distinct groups of ‘media repertoires’, distinguished by the level of their political-ideological diversity. This method has allowed us to see evidence of patterns that are broadly compatible with levels of news media polarization. Our results challenge the rather widespread narrative that people are nowadays enclosed in impervious ideological ‘bubbles’ or ‘echo chambers’ by showing that the majority of respondents in our sample are characterized by ‘open’ media diets, being exposed to content from across the ideological and political spectrum. However, our data also show that the countries with the highest proportion of ‘closed’ news diets are those with the deepest levels of media polarization. In Hungary, more than one-third of people fall within one of the ‘closed’ repertoires (mostly the anti-government, liberal one); while, in Serbia, the share of people getting their news from politically and ideologically homogeneous sources reaches nearly half of the population, with most of those audiences relying almost exclusively on pro-government, conservative media, thereby illustrating the hegemonic scope of the illiberal public sphere. On the other hand, the proportion of ‘closed’ media repertoires in Poland is 20% and only 16.6% in the Czech Republic. Furthermore, the relatively lower extent of media polarization in the latter country is complemented by a well-sized ‘balanced’ news media repertoire (31% of audiences), making the Czech Republic an outlier in the sample.

Through a subsequent analysis, we have remarked significant associations between the respondents’ political attitudes and the political and ideological profiles of the media which they regularly consume, as well as (though less unequivocally) their voting behaviour. Generally speaking, the data confirm that the more people are exposed to conservative and pro-government news sources, the less likely they are to display liberal stances on culturally polarizing issues (i.e. same-sex marriage or immigration), and the more likely they are to vote for parties associated with illiberalism. Conversely, people displaying the most liberal cultural attitudes and the highest respect for democratic institutions are those whose news diets tend to be composed predominantly of liberal, anti-government sources. This sample is also more likely to vote against the main governing party which, in all four countries, at the time of data collection, was the one enabling the rise of the illiberal public sphere.

We acknowledge that the results presented above are certainly not a proof of causality. Given the limitations of our data, particularly the cross-sectional nature of the survey and the selective character of the news media sample, it is not possible to prove whether exposure to news media with a particular editorial bias determines people’s political attitudes and electoral behaviour, or whether pre-existing political and ideological leanings determine people’s news choices. Nevertheless, it is safe to assume that both variables are mutually reinforcing each other and jointly affecting polarization, further deepening the divides in the media landscape and emptying its political-ideological centre. With regard to people’s attitudes, this process is more likely to have an asymmetric rather than a symmetric effect, especially in countries with an advanced illiberal public sphere, as demonstrated by the case of Hungary, where only the people with closed liberal and anti-government news diets hold culturally liberal views, while all the other news repertoire groups (composing 75% of the total audience according to our survey) are, on average, closer to the illiberal end of the attitudinal spectrum. One explanation for this could lie in the fact that, as the illiberal public sphere progresses and captures ever more mainstream media brands, people who generally prefer liberal sources but are also open to consuming news from the other side of the political-ideological spectrum (a group that, in Hungary, amounts to over 50% of the population) will more frequently be exposed to the illiberal views actively pushed by such mainstream brands, thereby becoming more susceptible to shifting their own opinions on culturally divisive issues. This, in other words, is what can be described as the process of normalization or mainstreaming of illiberalism, something that we examine in more detail later in the book. In the following Chap. 4, we examine how the growing polarization of media landscapes that accompanies the advance of the illiberal public sphere affects the levels and nature of media trust.